The ABINIT Users Mailing List ( CLOSED )

[文沈 <wshen02@gmail.com>]

I changed my shiftk as follows
 nshiftk 1
 shiftk  0.0 0.0 0.5
ngkpt 8 8 8 
I know the number of ngkpt is not enough, but I think at least it should give some graph which have the more or less the same peaks, right?
Besides, how can I treat that as metallic system? what should I change except diemac?

Thanks a lot,. guys!

Wen

On Tue, Mar 10, 2009 at 1:55 AM, matthieu verstraete <matthieu.jean.verstraete@gmail.com> wrote:

your kpoints are completely inappropriate:
1) you need a much finer mesh for graph*, especially for dielectric response
2) the shiftk you use are for an fcc structure, not hexagonal. Please read the appropriate tutorials (3+4) carefully.
3) as PMA mentioned, for graphene (an almost metal) and graphite (because its gap is ridiculously small) you should be treating the system as metallic and including the K point in the samplint (3n x 3n mesh of kpoints). This last point is less crucial once you converge the mesh properly.

Matthieu

On Tue, Mar 10, 2009 at 8:14 AM, Anglade Pierre-Matthieu <anglade@gmail.com> wrote:

First, I've never done any optic calculation. So I'm not a specialist.
Now : The dielectric function is the response of your material to an
electric field. You know that metals and semiconductors have very
different response to this. Then if one considers only the k-point
where there is a gap in graphen I would expect a noticeable effect on
the dielectric function.

regards

PMA

On Mon, Mar 9, 2009 at 11:34 PM, 文沈 <wshen02@gmail.com> wrote:
> Does this point matter a lot? I am just using the normal sampling as follows
> nshiftk 4
>  shiftk  0.5 0.5 0.5
>          0.5 0.0 0.0
>          0.0 0.5 0.0
>          0.0 0.0 0.5
>  ngkpt 4 4 4
> could you tell me why the sampling of the k point will have a effect on the
> calculation result?
>
> yours
> Wen
>
> On Mon, Mar 9, 2009 at 12:52 PM, Anglade Pierre-Matthieu <anglade@gmail.com>
> wrote:
>>
>> Does your reciprocal space sampling include the point where conduction
>> bands crosses the valence bands ?
>>
>> On Mon, Mar 9, 2009 at 6:46 PM, 文沈 <wshen02@gmail.com> wrote:
>> > Dear Abinit users,
>> > I am currently using abinit to calculate the dielectric function of
>> > graphite
>> > and graphene. But the problem is whatever pseudopotential I use, I get
>> > very
>> > bad result which I mean I get nearly all 0 for the imaginary part. I've
>> > already made sure that my structures are right.
>> >
>> > So anybody who calculate the dielectric function for Carbon please give
>> > me
>> > some idea what is the problem.
>> > Thanks a lot !
>> >
>> > Yours
>> > Wen
>> >
>>
>>
>>
>> --
>> Pierre-Matthieu Anglade
>>
>
>

--
Pierre-Matthieu Anglade

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Matthieu Verstraete

European Theoretical Spectroscopy Facility (ETSF) 
Dpto. Fisica de Materiales, 
U. del Pais Vasco,            
Centro Joxe Mari Korta, Av. de Tolosa, 72,   Phone: +34-943018393
E-20018 Donostia-San Sebastian, Spain        Fax  : +34-943018390

Mail : matthieu.jean.verstraete@gmail.com
http://www-users.york.ac.uk/~mjv500